正交异性薄板最大应力值的灰色神经网络算法

把灰色系统理论和神经网络模型结合起来，建立了相应的灰色神经网络模型。对有限宽正交异性薄板应力集中问题进行分析，根据其应力集中处附近3点的应变值，来确定最大的应力（应变）值。结果表明，该方法使用方便，数据吻合性较好，结果可靠。

Grey-neural Network Method to Determine Maximum Stress of Orthotropic Thin Plate

Stress concentration is very common in engineering. It is the major factor which destroys building block. Because of restrictions of lots of factors, the maximum stress or strain in the region of stress concentration couldn't be obtained directly. Those occur frequently in multiplex material building block. In this paper, an adequate model was established by combining grey model and neural network model. According to three points strains in the region of stress concentration of orthotropic thin plat, the maximum strain is obtained by training the network. Analysis shows that the method is efficient and reliable, and Neural network has good self study function, it is a kind of effective research tool for solving nonlinear systematic forecast problem.